In traditional methods of creating audio with embedded messages, the embedded messages are typically played back at native frequencies and speeds, which limits the number of embedded messages that is present in a given audio recording. Accordingly, there is a need for a more efficient and effective way to generate audio data containing embedded messages on a “per time” and/or a “per file size” basis.